The mechanism of genomic imprinting is unknown. Our previous studies have allowed us to propose a novel model for the mechanism of imprinting. This model for the function of the 15q11-q13 imprinting centre (IC) can be tested using novel transgenic techniques developed by this group. The IC is the region required for normal establishment of imprinting throughout the surrounding 2 Mb imprinted domain. We have demonstrated the unexpected characteristic of heterochromatin-forming DNA at the IC. This led to the hypothesis that imprinting occurs by forming heterochromatin at these sequences in oogenesis and by disrupting this organisation in spermatogenesis. This correlates with the heterochromatin-like characteristics of the maternally-inherited chromosome, which is more methylated, less nuclease accessible and later replicating than the paternal chromosome. This model also implies a simple mechanism for imprint spreading over large distances, by suppressive position-effect, as described in Drosophila and yeast. The first test of this model will involve the use of a MATH20 inducible construct in cell lines to see whether its expression alters imprinting of 15q11-q13. MATH20 is an artificial protein that has an extremely high affinity for MAR sequences, and has been shown to bind to heterochromatin-forming DNA and disrupt its functional effects. We will be able to analyse whether an agent disruptive of heterochromatin organisation interferes with the maintenance of imprinting of 15q11-q13, allowing a preliminary test of our model. Definitive testing of the model requires transgenic analysis in mice. Testing models of epigenetic regulation transgenically is difficult because of the tendency of transgenes to be influenced by epigenetic artefacts. To eliminate these artefacts, we have developed a novel cloning vector to allow the targeted integration of large single-copy transgenes. We will test the imprinting of IC and ZNF127-containing transgenes to see whether they behave as predicted by the heterochromatin-switching model. In addition, we will be able to test the imprint spreading prediction by targeting the transgenes to the mouse beta-globin locus and testing whether the beta-globin genes become subject to suppressive position-effect. Both the investigation of the heterochromatin-switching model and the development of the novel transgenic technology will provide the basis for further projects to test the roles of a range of factors in epigenetic processes such as genomic imprinting.